Buildings located in seismic risk areas are susceptible to damage, so evaluation of the seismic performance of building structures is essential for accurate seismic design. Seismic Damage Assessment (SDA) is a difficult task for individual building structures dealing with FiniteElement Analysis (FE) and simulation techniques are time-consuming due to the complex model. Moreover, computational techniques based on FE with Soil-Structure Interaction (SSI) for earthquake damage assessment of buildings require high computational efforts to build the database for the development of the area-based prediction model. Therefore, this study provides a framework for an artificial neural network (ANN)-based model as a reliable alternative to developing a rapid decision-making tool for building SDA. To establish the SDAbased ANN model, three inputs, including seismic, building, and soil were selected as the main parameters, and seismic responses with soil-structure interaction were generated using a multi-step analysis process proposed. The results obtained demonstrate the high accuracy of the SDA-Net model proposed as a predictive approach and rapid decision support tool for structures with SSI impacts as a predisaster management approach for the assessment of damage and a handy tool for safe and durable structures against natural disasters.